IMDHMIControls LibraryShop main store

Selection guide

How to select a CODESYS fieldbus master for industrial Ethernet

Industrial networks Published 9 min read

Short answer

Select the master around the device, not the connector: the 750-340 is a PROFINET IO coupler, and a PROFIBUS station needs a different coupler. Confirm three matches — device protocol, runtime master package, supported network adapter — then size the bus cycle from the process budget; the worked example clears a 10 ms sensor-to-actuator budget at a 2 ms cycle and fails it at 4 ms. Bench-test a laptop NIC only if the runtime documents it.

On this page
  1. Does the WAGO 750-340 coupler speak PROFINET or PROFIBUS?
  2. Does an Ethernet port alone make a device PROFINET-capable?
  3. Which master package does the CODESYS runtime need?
  4. Can a standard laptop Ethernet port run the fieldbus master?
  5. Which protocol meets the update time the process needs?
  6. What do you check on the station before it runs the machine?
  7. What do you record when the master still will not start?
  8. So which master do you actually buy?
  9. Questions asked

Match the master to the device, not to the connector. Before any interface hardware is bought, three things have to agree: the remote I/O device speaks the protocol you configure, the installed runtime includes the matching master package, and the runtime's real-time driver supports the adapter you plug in. The checks below run in the order a selection actually happens — identify the coupler, apply the IEC profile families, confirm the runtime and its packages, test the adapter, then size the bus cycle from the process update budget. Every step ends in a decision, and the first one is made at the coupler's nameplate, not at the network switch or the laptop bag.

Does the WAGO 750-340 coupler speak PROFINET or PROFIBUS? #

The 750-340 is a PROFINET IO fieldbus coupler: the product listing places it in the PROFINET IO family, and the coupler auto-configures a local process image covering the 750-series I/O modules plugged behind it. PROFINET and PROFIBUS couplers in the same 750 family are separate part numbers built around different network controllers, so a station that runs PROFIBUS today carries a PROFIBUS coupler no matter which patch cords sit in the cabinet drawer. Read the part number off the nameplate and the configured bus out of the running project before anything else, then let the table decide.

What the nameplate shows What the device speaks Decision
750-340, PROFINET IO PROFINET over Ethernet Configure a PROFINET master and keep the 750 I/O modules unchanged
PROFIBUS coupler marking PROFIBUS DP over RS-485 Keep a PROFIBUS master, or replace the coupler to change buses
EtherCAT marking EtherCAT over Ethernet EtherCAT master only; the RJ45 jack does not confer membership

Decision: if the installed coupler is the 750-340, PROFINET is the protocol and the PROFIBUS question is closed; if the station really runs a PROFIBUS coupler and must stay untouched, the master is PROFIBUS and the Ethernet-port discussion ends there. Nothing about a cable, a switch or a laptop changes either outcome.

Does an Ethernet port alone make a device PROFINET-capable? #

An RJ45 jack settles only the physical layer: IEC 61158 and IEC 61784 sort industrial networking into communication profile families — PROFINET family 3, EtherNet/IP family 2, EtherCAT family 12 — and membership comes from the firmware, not from the connector.

IEC 61158 defines the data-link services each industrial network provides, and IEC 61784 collects them into those profile families, each maintained by its own protocol organization. A device joins a family by implementing that family's application behavior: a PROFINET device takes a name and an IP address, then exchanges cyclic I/O and diagnostics as the family defines them; an EtherCAT slave processes telegrams on the fly as the frame passes through; an EtherNet/IP device exposes CIP objects. PROFIBUS DP and PROFINET share profile family 3, yet PROFIBUS runs on RS-485 and PROFINET on Ethernet, so sharing a family is not interoperability.

For a selection the consequence is simple: treat every statement that a device has an Ethernet port as a physical-layer fact only, and let the datasheet's protocol section decide. That is why the coupler check comes first, and why a PROFINET master cannot reach a coupler built for another bus.

Which master package does the CODESYS runtime need? #

CODESYS documents support for EtherCAT, PROFINET and EtherNet/IP inside its IEC 61131-3 programming system, so the protocol choice is rarely blocked at the tool level — the real check is that your installed runtime includes the matching master stack and its license.

Fieldbus masters ship as add-on packages, so make the check concrete: on the engineering machine, confirm that the installed runtime (a CODESYS SP RTE V3.3 install is checked as its exact build) lists the master package for the protocol the coupler needs and that the license covers it. Supported adapters are stated per runtime version, so read the release notes and the runtime's communication documentation instead of assuming the previous version's adapter list still holds. If the package is missing, the missing piece is a package and a license, not a cable.

On Windows-based runtimes such as Control RTE, real-time fieldbus operation needs the runtime's own network driver bound to a specific adapter through the runtime's network driver setup. After the binding, that adapter stops serving ordinary Windows traffic, which is why the setup step exists. For the messaging side of EtherNet/IP — reading and writing tags in a Logix controller from CODESYS — the method is covered in the article on reading and writing Logix tags from CODESYS with CIP GenericService.

Can a standard laptop Ethernet port run the fieldbus master? #

Yes for bench development, but only when the runtime documents that adapter as supported by its real-time driver — an ordinary TCP/IP link light proves the cable, not that the master's cyclic traffic is compatible with the NIC.

Real-time fieldbus traffic puts timing demands on the interface that ordinary networking never tests: the driver has to hand frames to the stack when the master cycle needs them, not when the operating system gets around to it. The test is short. Open the runtime's network driver setup and check whether the built-in adapter is recognized and can be bound; if it binds, bench-test one coupler and watch the master status for adapter and update errors; if it does not bind, that port is out of the selection.

A dedicated real-time Ethernet interface card — the CIFX 90-RE is one example of this class — becomes an option only when the runtime documentation lists it as a supported interface for your protocol and version; the card name alone establishes nothing. If the master will live on its own machine in the cabinet rather than a development laptop, host the runtime on an industrial mini PC specified for continuous duty instead of repurposing office hardware. What the card costs is a purchasing question; whether one is needed at all is decided by the binding test above.

Which protocol meets the update time the process needs? #

Size the protocol from the device you must keep and the update time the process needs: in the worked example below, a 10 ms sensor-to-actuator budget clears at a 2 ms bus cycle and fails at 4 ms, so the arithmetic decides.

A sensor bit reaches an actuator through a fixed chain: the coupler's update time, one bus cycle to the master, the PLC task, one bus cycle back, and the output device's update time. Estimate the worst case as t_task + 2 × t_bus + t_update, where the times are in milliseconds and the formula assumes the sample misses one cycle in each direction. The settings below are illustrative, not protocol minima; your configured values replace them.

  • Inputs: PLC task period 2 ms; coupler update time 4 ms (case A) or 2 ms (case B); bus cycle 4 ms (case A) or 2 ms (case B); process budget 10 ms sensor-to-actuator.
  • Case A: 2 ms + (2 × 4 ms) + 4 ms = 14 ms, which misses the 10 ms budget by 4 ms.
  • Case B: 2 ms + (2 × 2 ms) + 2 ms = 8 ms, which meets the budget with 2 ms to spare.
  • Decision: this budget needs the 2 ms cycle with the 2 ms coupler update, and every doubling of task period or update time moves the sum by the same amount.
Protocol IEC 61784 family Device side must be Runtime needs Choose when
PROFINET Family 3 PROFINET device firmware, described by GSDML PROFINET master package plus a supported real-time adapter The installed station is PROFINET-ready, as the 750-340 is
EtherNet/IP Family 2 EtherNet/IP adapter, described by EDS Scanner package The device ecosystem is CIP-based and explicit messaging matters
EtherCAT Family 12 EtherCAT slave, described by ESI EtherCAT master package The budget demands cycles the installed devices cannot reach

Decision for this project: all three protocols clear a 10 ms budget at ordinary settings, so speed does not pick the protocol — the installed 750 station does. PROFINET wins because the coupler speaks it; EtherCAT becomes the answer only when the devices on the wire are EtherCAT slaves, because moving protocol means moving devices.

What do you check on the station before it runs the machine? #

Bring the station up in layers — the runtime binds the adapter, the master starts without configuration errors, the coupler appears under the configured protocol, then I/O data updates in the controller — and keep machine outputs disabled until the mapping is proven.

  1. Bind the adapter in the runtime's network driver setup and confirm the master starts with no adapter error.
  2. Load the device description from the coupler manufacturer and match it against the nameplate part number.
  3. Set the address the protocol uses — name and IP address for PROFINET — and confirm the master reaches the device.
  4. Prove the process image: observe one input and one output in the controller before anything moves.
  5. Read the coupler's own diagnostic indicators for protocol state; a switch's link lights say nothing about the fieldbus.

For the station-side steps in order, the remote I/O commissioning checklist for a modular station walks the coupler-to-controller bring-up, and the outputs-disabled line above stays in force the whole time.

What do you record when the master still will not start? #

Capture five facts before asking anyone: the exact runtime version and build, the adapter model and driver version the runtime bound, the coupler part number with its configured bus, the protocol and the device description file loaded, and the master's own status output.

Stop hardware selection when any of three things happens: the part number or protocol marking is unclear, the runtime cannot bind any supported interface, or the device stays undiscovered after the protocol and configuration checks above. Escalate with the five facts to the manufacturer's official support channel — the runtime vendor for master and adapter errors, the coupler manufacturer for device-side errors — rather than retrying different cables until something changes.

Keep the diagnostics protocol-specific once the bus is up. If the network turns out to be EtherCAT, its master can keep xError false after a cable pull, as the article on why a CODESYS EtherCAT master's xError stays false on cable pull explains; a quiet error flag is not proof of a healthy link.

So which master do you actually buy? #

Buy the master the installed devices force, in this order: a PROFINET master if the coupler is the 750-340, a PROFIBUS master if the station is a PROFIBUS coupler that must stay, and interface hardware only after the runtime documents the adapter.

  • Installed coupler is the 750-340: configure a PROFINET master, keep the 750 I/O modules, run the 2 ms class cycle that produced the 8 ms result in the example, bench-test the laptop NIC only if the binding test passes, and put a documented interface in the panel.
  • Station is a PROFIBUS coupler that must stay: buy a PROFIBUS master; no Ethernet interface hardware belongs in this selection, and PROFINET re-enters only when the coupler is replaced.
  • Budget tighter than the installed devices support: choose EtherCAT devices and an EtherCAT master together, because the protocol and the device base move as one purchase.
  • Laptop port: a bench tool when documented, never the untested interface in a running panel.

Every case resolves at the same point: the coupler's nameplate decides the protocol, the runtime decides the package, and the driver decides the adapter — in that order.

Questions asked

Can one CODESYS runtime run PROFINET and EtherNet/IP masters at the same time?

Yes, when the installed runtime includes both master packages, the license covers both stacks, and each master has its own interface or port assignment. The practical limits are the runtime version and the number of real-time Ethernet interfaces the host exposes. Check the runtime's communication documentation for the exact version before planning a mixed network, and keep each protocol on its own adapter so diagnostics stay separated.

Is the EtherNet/IP Adapter SL a scanner or a device?

It is the device side. The published description says the add-on turns a CODESYS-compatible PLC with standard Ethernet ports into an EtherNet/IP adapter, which is the node a scanner connects to. A scanner role for controlling other EtherNet/IP devices is a separate package, so confirm which role your application needs before you install or license anything.

Can I take GSDML, EDS or ESI files from a file-sharing site?

No. Get description files from the device manufacturer or its authorized distributors. PROFINET devices use GSDML, EtherNet/IP devices use EDS, and EtherCAT devices use ESI. A description file that does not match the device's firmware produces mismatches that look like network faults, and modified files put unverifiable configuration into a live project, so the vendor source is part of the engineering decision, not paperwork.

What changes if the master runs on an embedded controller instead of a PC?

Adapter choice disappears: the real-time Ethernet interface is built into the controller hardware, and the manufacturer documents which protocol that hardware serves. TI, for example, publishes an application note on its AM6x processors running CODESYS that includes the hardware required to set up an EtherCAT network. Read that per-target documentation before sizing the network, because a PC runtime's supported-adapter list does not transfer to an embedded part.

Keep reading

  • Troubleshooting

    Why CODESYS EtherCAT master xError stays false on cable pull

    `EtherCAT_Master.xError` stays FALSE because the master does not latch slave loss into the application variables when restart handling is off; the engineering tool reads device state directly, so it still shows the error. Tick `Automatically restart slaves` on the master and `Enable Diagnosis for Devices` for the extended instance, then pull-test: the loss reaches the program and the slave rejoins when the cable returns.

    Industrial networks

  • Selection guide

    Choosing PLC to SCADA communication over long distances

    For PLC-to-SCADA communication over long distance, choose the transport by reach and connection behavior: fiber Ethernet for a continuous long link (one span rated 2 km), RS-485 where serial fits, extenders or radio on existing pairs, dial-up or leased-line modems on phone service. Direct copper Ethernet fails at 100 m per segment; the SCADA driver or OPC service, not the PLC's ASCII modem commands, carries the protocol.

    Industrial networks

  • How-to

    CompactLogix MSG to SLC 5/04 through a DH+ gateway

    The CIP path in the CompactLogix MSG ends at the 1756-DHRIO module in the remote chassis. On the MSG configuration tab, change the communication method from CIP to DH+, then enter Channel A and the SLC 5/04 node address in the DH+ fields; the SLC is never part of the path itself.

    Industrial networks